'If the stamens are longer than the corolla they are exserted; if shorter, they are called included: when they all bend to one side, as in Amaryllis, they are declinate; if two out of four are shorter, they are didynamous; if four out of six are longest, they are tetradynamous.

The number of stamens is indicated by a Greek numeral prefixed to the word androus, which signifies male, thus:

The filament (Plate III.) (capillamentum, or pediculus, of some) is the part that supports the anther. It consists of a bundle of delicate woody tissue and spiral vessels, surrounded by cellular tissue, and is in all respects the same as the petiole of a leaf, of which it is a modification, except that its parts are more delicate. As the petiole is unessential to the leaf, so is the filament to the anther, it being frequently absent, or at least so strictly united to the sides of the calyx or corolla as to be undistinguishable. Its most common figure is filiform or cylindrical (Plate III. fig. 12, 13. 20, 21.), and it is almost always destitute of colour; but there are exceptions to both these characters. In Fuchsia, for instance, the filaments are red like the petals; in Adamia they are blue; in ŒŒnothera they are yellow; and a return to the foliaceous state of which they usually are a distinct modification is by no means rare. (Plate IV. fig. 6. 8.) Thus the filament in Canna is undistinguishable from petals except by its having an anther; in the same genus and its allies, and in all Zingiberaceæ, the

inner series of what seem to be petals are modifications of filaments (see Introduction to the Nat. Syst., ed. 2. p. 325.) : and this is a very common circumstance in sterile stamens.

The filament also varies in other respects: in Thalictrum it is thickest at the upper end, or clavate (Plate III. fig. 23.); in Mahernia geniculate (Plate III. fig. 25.), in Hirtella spiral, in Crambe bifurcate, in Anthericum bearded or stupose. In some plants the filaments are combined into a solid body called the columna, as in Stapelia, Stylidium (Plate IV. fig. 1, 2, 3.), Rafflesia, and others: this has in Orchidaceæ received from Richard the name of gynostemium.

Care must be taken not to confound the pedicel and single stamen of the naked male flowers of Euphorbia with a filament, as was done by all writers, until Brown detected the error. For modifications of filaments see Plates III. and IV.

The Anther (Theca of Grew; Capsula, Malpighi; Apex, Ray; Testiculus or Testis, Vaillant; Capitulum, Jungius; Sper matocystidium, Hedwig) is a body generally attached to the apex of the filament, composed of two parallel lobes or cells (thecæ, or coniothecæ, or loculi,) containing pollen, and united by the connective. It consists entirely of cellular tissue, with the exception sometimes of a bundle of very minute vascular tissue, which diverges on each side from the filament, and passes through that part of the anther from which the pollen has been incorrectly supposed to separate, and which is called the receptacle of the pollen by some, the trophopollen by Turpin, and the raphe by Link, but with greater propriety the septum of the anther. Its coat is called by Purkinje exothecium.

In the most common state of the anther the cells are parallel with each other (Plate III. fig. 14.), and open with two valves (Plate III. fig. 13. a), by a longitudinal fissure from the base to the apex; in Labiatæ and Scrophulariaceæ the cells diverge more or less at the base (Plate III. fig. 15, 18.), so as in some cases to assume the appearance of a one-celled horizontal anther, especially after they have burst. In Cucurbitace the lobes are very long and narrow, sinuous, and folded back upon themselves (Plate III. fig. 24.). In Salvia

the connective divides into two unequal portions, one of which supports a cell and the other is cell-less; in this case the connective has been called by Richard, distractile. Lacistema (Plate IV. fig. 7.) affords another instance of a divided connective. In many of the cases of excessive divergence of the cells the line of dehiscence of the anther is changed from longitudinal to vertical (Plate III. fig. 20. 17.), and has actually been supposed to be really transverse; an error which in most cases has arisen from not understanding the real structure of the anther. Some anthers, however, no doubt have cells that burst transversely, as Lemna, Alchemilla arvensis, Securinega, &c. (See Plate III. fig. 12. 16. 30.)

All anthers are not two-celled, their internal structure being subject to several modifications. It sometimes happens that the anther is four-celled, as in Tetratheca. In Epacris the two ordinary cells become confluent into one, and the anther is therefore one-celled. In Maranta and Canna only one cell is produced, the other being entirely suppressed. In most Amarantaceæ, and some other plants, the anther seems to be absolutely one-celled. (Plate IV. fig. 8.)

Of all these the four-celled anther is the type, and both the one-celled and two-celled are probably mere modifications of it, depending upon whether the septa which originally exist all remain complete, or are half absorbed, or wholly absorbed. Schleiden says he has found the anther before its bursting quadrilocular in more than one hundred families; amongst which may be named Graminaceæ, Cyperaceæ, Liliaceæ, Labiatæ, Borraginaceæ, Scrophulariaceæ, Compositæ, Apiaceæ, Ranunculaceae, with its allies, Rosacea (Juss.), and Leguminosa, which orders alone constitute almost one half of the entire vegetation of the globe. It has been often asserted, he says, that the anther could not originally be quadrilocular, because it opens by two fissures only; which is as much as to consider two rooms in a house as one, because they have not folding doors, but single doors placed close together. Properly speaking, every anther really opens with four fissures; they appear, however, only as two, because each pair lies at the side of the common septum.

Other deviations from the normal form of anther occur,

which are less easy to reconcile with the idea of a two-celled type. In some Lauracea the anther is divided into four cells, one placed above the other in pairs; in Ægiceras it consists of numerous little cavities; and in the singular genus Rafflesia the interior is separated into many cellules of irregular figure and position, described by Brown as "somewhat concentrical, longitudinal, the exterior ones becoming connivent towards the apex, sometimes confluent, and occasionally interrupted by transverse partitions." In these instances the septa may be understood to arise from portions of the cellular tissue of the anther remaining unconverted into pollen.

With regard to deviations from the usual mode of dehiscence, Brown observes (Linn. Trans. xiii. 214.), "that they are numerous: in some cases consisting either in the aperture being confined to a definite portion, generally the upper extremity of the longitudinal furrow, as in Dillenia and Solanum ; in the apex of each theca being produced beyond the receptacle of the pollen into a tube opening at top, as in several Ericaceae (Plate III. fig. 22.); or in the two thecæ being confluent at the apex, and bursting by a common foramen or tube, as in Tetratheca (see Plate IV. fig. 4.). In other cases a separation of determinate portions of the membrane takes place, either the whole length of the theca, as in Hamamelaceæ and Berberaceæ, or corresponding with its subdivisions, as in several Lauraceæ, or lastly, having no obvious relation to internal structure as in certain species of Rhizophora." In Lauraceae and Berberacea the anthers are technically said to burst by valves (Plate IV. fig. 10, 11.), that is to say, the dehiscence does not take place by a central line, but the whole face of the cell separates from the anther, and curls backwards, adhering to it only at the apex, to which it is, as it were, hinged.

In Rhizophora, above alluded to, the anther is said by Mr. Griffith to be compressed, with the edges anterior and posterior, and to open by the separation of a valve from the sides, when the pollen is seen lying in alveolar excavations, the upper portions of which may be traced on the inner face of the valves. (Trans. Med. & Phys. Soc. Calcutta.) Mr. Griffith rightly explains this singular structure as a modifica

N

tion of dehiscence, caused by the adhesion of the valves to each other at the usual line of fissure, and their separation from the connective. It is not therefore quite correct to say that the dehiscence of Rhizophora bears no obvious relation to internal structure; it is of the same nature as that of Lauraceæ, Berberaceæ, &c., and is analogous to the disarticulation of the valves of Brassicaceous plants.

The cells of the anther have frequently little appendages, as in different species of Erica, when they resemble setæ, aristæ, or crests. (Plate III. fig. 29.)

The anthers are attached to the filament either by their base, when they are called innate (Plate III. fig. 27. 21. 23.), or by their back, when they are adnate (Plate III. fig. 13.), or by a single point of the connective from which they lightly swing in the latter case they are said to be versatile. This form is common to all true Grasses.

When the line of dehiscence is towards the pistil, the anthers are called by Brown antica, but by other botanists introrsæ, or turned inwards: when the line is towards the petals, they are said by Brown to be postice, and by other botanists to be extrorsa, or turned outwards.

The connective is usually continuous with the filament, and terminates just at the apex of the anther; but in some plants, as Compositæ, it is articulated with the filament (Plate IV. fig. 5.). In others it is lengthened far beyond the apex (Plate IV. fig. 6. 9.), now into a kind of crest, as in many Zingiberaceae; now into a sort of horn, as in Asclepiadaceæ ; now into a kind of secreting cup-like body articulated with the apex, as in Adenostemon. Very frequently it is enlarged in various ways. For cases of this kind, see Plates III. and IV. Its being sometimes two-lobed, or forked, has been already noticed (Plate IV. fig. 7.). The lining of the anther has received particular illustration from Purkinje, who calls it endothecium, and who has found that it consists of that very remarkable kind of tissue, which has been already described under the name of fibro-cellular. According to that botanist the forms of this tissue are extremely variable, the vesicles being sometimes oblong, sometimes round, frequently cylindrical, usually fully developed, or, in some cases, merely